[FRIAM] 100 billion neurons

[Jochen Fromm]

Interesting remark, but I don't think it really works this way. 
It is not an infinite ensemble of an infinite number of analog events. 
A neuron fires or not - a boolean event - and spikes are certainly 
discrete events. The ion channels, the gradients of ions, and all 
the chemical substances are only the "hardware" of the brain. One 
could compare it to transistors, wires, etc. If the genes could 
produce transistors instead of proteins, they would perhaps use 
digital circuits. However, the interesting part seems to be the 
software, esp. the code which is used (if there is any). There 
are of course at least four different levels of modelling, 
from boolean networks and sigmoid networks to spiking networks, 
see Fig. 3 in http://www.vs.uni-kassel.de/~fromm/Articles/LI.pdf

-J.
 

-----Original Message-----
From: friam-bounces at redfish.com [mailto:friam-bounces at redfish.com] On Behalf
Of doug
Sent: Sunday, July 09, 2006 2:24 AM
To: 'The Friday Morning Applied Complexity Coffee Group'
Subject: Re: [FRIAM] 100 billion neurons

Comment: this implies a discrete ensemble of discrete events. But isn't each
neuron's likelihood of firing dependent on the solution in which it sits,
the gradients of ions, and proximities to tier multiple firing neurons?

In which case the brain is an infinite ensemble of an infinity of analog
events.

doug